Jump to content

Sabermetrics, My Take

brianod

By brianod
in Orioles Talk

Recommended Posts

ledzepp8 Hangout Grand Counsel

ledzepp8

Posted
Posted
1- No one denies this. The "eye test" also has issues with small sample sized, I would say more issues.

2- Who are these "many people"?

3- It is also impossible to measure then with the "eye test". The very word intangible should be a clue that it isn't something easily measured.

4- I think analytics can be of great use when looking at young talent. For instance Manny was very young for his level so his bat projected better then the raw numbers indicated. Same with Schoop last season, sabermetrics provided a framework to see how he might project forward. Whereas someone just looking at his actual performance might think he would could never be a ML hitter.

5- Yes, we all know this. That would be why you read people saying things like "It takes two seasons for defensive metrics to stabilize".

I think there is a lot more discussion available when stats are used then when folks make blanket statements without supporting arguments then refuse to budge from them.

I don't understand why folks rebel against more knowledge. How can you not want to know more? Yes you have to pay attention and weigh the value of what you ingest but it sure beats ignorance.

Because obviously you should base all of your positions and arguments off of intuition and anecdotal evidence.

Link to comment
Share on other sites
  • Replies 206
  • Created
  • Last Reply
weams Hangout Grand Counsel

weams

Posted
Posted
Because obviously you should base all of your positions and arguments off of intuition and anecdotal evidence.

Well a friend of mine said he heard on talk radio that there was a difference in the clubhouse without Nick. I could only speculate that is why Steve Pearce is not hitting well. He thinks we might only have 78 wins also!

Link to comment
Share on other sites
ledzepp8 Hangout Grand Counsel

ledzepp8

Posted
Posted
No, I'm saying when you summarize twenty different stats into value, you lose a lot. I think that's by definition, some may disagree. For example, Altuve is a great player but is he worth the same WAR in every stadium? Isn't that a supposition of the WAR summary? In my opinion, guys that play for the Orioles need a little pop. They play 81 games in a bandbox. A player who plays above average defense and hits for power means more to the Orioles than some teams. If I'm wrong, I'll eat my words. Also, your division matters. Where do you play the most, outside of your home stadium? Does WAR account for that? I could go on and on, but imagine some on this board wouldn't want that:)

I'm obviously far from an expert on sabermetrics, but I'm pretty sure that park factors are taken into account with WAR.

Link to comment
Share on other sites
ledzepp8 Hangout Grand Counsel

ledzepp8

Posted
Posted
So at what point isn't there a small sample size? How can player who played as much as Lough did last year be properly judged in relation to WAR?

I also think that OPS+ and ERA+ make some sense but they don't tell me what a player would do on a different team. I don't think if Player A plays in a hitters park and has an OPS+ of 123 and Player B plays in a pitchers park and has the same numbers that means they are the same player. At the end of the day If I am a GM it matters how that player performs for my team.

Last year WAR told us the Orioles pitchers were practically the same as 2014. Because WAR says so I am not suppose to believe my eyes?

OPS+ is supposed to be park adjusted...so yeah it's not saying they're the same player but it's saying they were similar in regards to OPS+.

Link to comment
Share on other sites
ledzepp8 Hangout Grand Counsel

ledzepp8

Posted
Posted

Forgive me if I'm looking at this wrong. Baseball Reference says that replacement level for a team is 48 wins. 2014 Orioles were worth a cumulative rWAR of 45.9. Add that to the 48 wins and you get 93.9 wins. Their Pythagorean W/L was 94-68. They went 96-66. It would seem like WAR is doing a pretty good job of judging the worth of that collection of players. Is it perfect? Of course not. But what is? It's certainly lends more accuracy than just looking at the team and saying, yeah they look pretty good.

Link to comment
Share on other sites
brianod All-Star

brianod

Posted
  • Author
Posted
Forgive me if I'm looking at this wrong. Baseball Reference says that replacement level for a team is 48 wins. 2014 Orioles were worth a cumulative rWAR of 45.9. Add that to the 48 wins and you get 93.9 wins. Their Pythagorean W/L was 94-68. They went 96-66. It would seem like WAR is doing a pretty good job of judging the worth of that collection of players. Is it perfect? Of course not. But what is? It's certainly lends more accuracy than just looking at the team and saying, yeah they look pretty good.

That is my entire point in this thread, is it perfect, nothing is. I'm simply saying that too much reliance can be placed on it, not that it isn't a tool that should be used.

Link to comment
Share on other sites
weams Hangout Grand Counsel

weams

Posted
Posted
That is my entire point in this thread, is it perfect, nothing is. I'm simply saying that too much reliance can be placed on it, not that it isn't a tool that should be used.

No one. No one. Says that the tools can't be wrong in a small sample. No one says J.J. can't get that hit. But math says 75 percent of the time, he won't. Over the long run.

Link to comment
Share on other sites
brianod All-Star

brianod

Posted
  • Author
Posted

To further some up and be done with this, player evaluation is an art, not a science. Human performance can't be judged entirely by statistics. The advanced metrics of today should be used in conjunction with human judgment. Simply saying that someone has more WAR doesn't win the argument. DD and Buck make big bucks along with their staffs and we should rely on their experience and judgment along with advanced metrics in making personnel decisions. I don't think this is all that controversial to say and do believe it's true. Let's go out and get a win today gentlemen, I sense a little positive momentum building after last night.

Link to comment
Share on other sites
Can_of_corn Hangout Grand Counsel

Can_of_corn

Posted
Posted
That is my entire point in this thread, is it perfect, nothing is. I'm simply saying that too much reliance can be placed on it, not that it isn't a tool that should be used.

Looking back on your first post, you were saying a good bit more then that.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Go to topic listing

  • Posts

    • deward
      David Rubenstein 2024

      By deward · Posted

      I'm looking at the Rays record over the past decade and I'm a bit confused over your definition of "have not won much". However, my argument is simple - teams that don't invest in payroll don't win championships in baseball. Over the past 21 years (I was going to do 20, but added the extra to include Florida's win in 2003), the average payroll position relative to the league of the eventual WS champ was 9th (see below) Year WS Champ OD Payroll rank 2003 FLA 25 2004 BOS 2 2005 CWS 13 2006 STL 11 2007 BOS 2 2008 PHI 12 2009 NYY 1 2010 SF 10 2011 STL 11 2012 SF 8 2013 BOS 4 2014 SF 7 2015 KC 16 2016 CHC 14 2017 HOU 18 2018 BOS 1 2019 WAS 7 2020 LAD 2 2021 ATL 13 2022 HOU 11 2023 TEX 9   Only three of these teams had payrolls in the bottom half of league: Florida in 2003 had by far the lowest payroll, at 25th. Their payroll climbed to as high as 18th over the next couple of years, but they couldn't maintain their success and haven't made the playoffs since, outside of the COVID season. KC in 2015 had the 16th lowest payroll, barely below the median payroll for the year. They haven't been back to the playoffs since. Houston in 2017 had the 18th lowest payroll. This was their big breakthrough year after their tanking/rebuild, and they haven't been lower than 11th since, and as high as 4th.   The trend is obvious. After the Marlins' miracle run in 2003, no team has won the WS with a payroll lower than 18th, and that team (Houston) is an obvious outlier as they were in the basically the same spot as the O's now (on the upswing from a full tear-down). While KC and Florida both had years where everything came together perfectly, they were unable to sustain their momentum. The O's were 23rd in payroll on Opening Day, and the current roster is good enough to win a championship, but history suggests they'll go the way of Florida and Kansas City if Rubenstein isn't willing to invest in the payroll. Consistently letting the talent drain out of your organization because you aren't willing to pay them won't lead to multiple championships and it won't keep fans engaged.
    • OriolesUpAndDowns
      Cal thinks Holiday moved to fast through the system

      By OriolesUpAndDowns · Posted

      Correlation vs. Causation: The study states that there is a correlation between 1RM (one-rep max) squat and performance metrics like the 10-yard split and 40-yard dash times. However, correlation does not imply causation. Just because these variables are correlated does not mean that one causes the other.   
    • Roy Firestone
      Baumann DFA

      By Roy Firestone · Posted

      I think he's claimed and right away.
    • baltfan
      Cal thinks Holiday moved to fast through the system

      By baltfan · Posted

      His Dad is a big boy and obvious lifter.  Some guys are naturally good squatters.  That’s an impressive number for Holliday, but who knows exactly how legit the actual squat is.  You can find high school kids at nearly every high school about his size with similar squat numbers.  Granted, they will be some of the strongest pound for pound kids, but it’s not a crazy number. 
    • Hallas
      Cal thinks Holiday moved to fast through the system

      By Hallas · Posted

      Fair point, but these outliers are squatting like 40% more than Holliday is.  There are likely plenty of guys in the NCAA that can hit Holliday's numbers, and are in his age range.   There were like 50+ guys at this year's combine with a 10 yard split of 1.5 or better or 40 times of 4.5 or better, and 1rm squat correlates with both 10 yard split and 40 time, such that if you hit those numbers in 10 or 40, then your mean 1rm squat would be predicted to be about 2.5x body weight, which would be right around 500 lbs for a 200 lb guy.  If you want to call NFL caliber college guys outliers, I guess?  But Holliday is an MLB caliber guy, so he still fits.   Elite athletes just aren't normal, man.
    • Moose Milligan
      Baumann DFA

      By Moose Milligan · Posted

      The fact that they put this on the 1991 Donruss, perhaps the ultimate peak of the junk wax era, makes this even funnier to me.  If toilet paper ever shot up in price I wouldn't care because I've got heaps of '91 Donruss floating around. This is very true, which makes me thinks he'll get claimed somewhere else.
    • Moose Milligan
      Cedric Mullins 2024

      By Moose Milligan · Posted

      It's like the ghost of Chris Davis found him and has taken over his body.   I know bat speed isn't everything, but with the bat speed info that dropped, I took a look.  His average is 70.9 which isn't terribly slow, it's actually pretty average.   For whatever reason, he just can't square anything up.  I hope he fixes it but it's really tough to watch right now.

  • Popular Contributors

×
×
  • Create New...